Two of 346 archive pairs keep Hill mass proxy ratios above two within three listed error widths under a 1% planet/star mass limit

We test how listed orbital sizes change a circular Hill mass proxy relative to a period-based estimate. The NASA Exoplanet Archive snapshot of 22 October 2026 supplies 958 adjacent transiting planet pairs in 639 systems with one listed star. At listed central values, 12 pairs have ratios above two and five have ratios below one half. One pair has reverse orbital size order. We then let both periods and both sizes vary independently within three listed error widths, with an added allowance for display rounding. We allow each planet to have up to 1% of its star's mass in the Kepler conversion. This is a conditional mass limit. The error test includes 346 pairs; 612 lack required errors or have invalid or zero errors. Two pairs keep the proxy ratio above two throughout the tested boxes. HIP 41378 e/f has a central ratio of 33.5002 and a range of 2.45747 to 184.90954. K2-285 b/c has a central ratio of 3.76810 and a range of 2.09053 to 6.03750. Neither remains outside the review limits at five listed error widths. Earlier work already found orbit value conflicts in these systems. The new contribution measures their effect on the Hill mass proxy. It supplies a gap atlas, 21 source-pair controls, exact snapshots, code, figures, and reproduction instructions. A separate 60-digit calculation confirms all 958 central decisions and all 5,166 tested interval decisions. The ranges are not probability intervals. The proxy is not a measured planet mass or a stability proof. The result supplies review targets before archive values enter a joint orbit model.

exoplanets · orbital spacing · mutual Hill radius · catalogue validation · NASA Exoplanet Archive · interval bounds · reproducible data